Adsorption of Safranin (Cationic) Dye from Water by Bambusa tulda : Characterization and ANN Modeling

Abstract

The present research is based on the removal of safranin (cationic) dye, an emerging contaminant from wastewater by sodium bicarbonate modified Bambusa tulda (variety of bamboo and a bioadsorbent) using adsorption technology. Characterization study of sodium bicarbonate-treated bamboo was done through scanning electron microscope analysis and it was found that after treatment with sodium bicarbonate, pore structure of B. tulda gets widened increasing active pore cavity in the adsorbent resulting in larger surface area for higher adsorption. Energy dispersive X-ray analysis showed the presence of Na, Cl, etc., in safranin dye adsorbed adsorbent; in Fourier transform infrared spectroscopy analysis presence of Hydroxyl and Carboxyl group helps largely for adsorption of safranin dye. The batch study experiment gave optimum results for safranin dye at pH 7; rpm 200, temperature 298 K, with an initial concentration of 50 mg/L and a dose of 10 g/L with maximum adsorption capacity of 32.26 mg/g, followed by pseudo second order rate equation and a removal rate of 98.15% for 60 min of time duration. The activation energy was 17.376 kJ/mol. Two algorithms are used in artificial neural network (ANN), that is, Levenberg–Marquardt (LM) and Bayesian Regularization algorithm with different nodes. LM algorithm with three hidden nodes has higher R² value with the lesser value of statistical error. Time, dose, and initial concentration can be considered for t-test. Intensive research has been carried out on the modified B. tulda along with characteristic study and modeling analysis by ANN and regression analysis Model Study and Error analysis. Taking considerations of all experimental results and model analysis, it is concluded that sodium bicarbonate treated B. tulda is a good, low cost, and effective adsorbent in comparison to commercially activated carbon for the removal of safranin (cationic) dye from wastewater.

Introduction

Wastewater treatment is a challenging objective in our modern world, and most of the developing countries are facing this problem. The waste liquid effluents, which are coming out of the industries like textiles, paper, leather, paint, cosmetics, pharmaceuticals, dyeing, wools, and cotton industries, are discharged into the water bodies without proper treatment or partial treatment, and it poses a great threat to ecosystem and environment (Mittal et al., 2010a). Dyes are one such pollutant which comes out through such industries. Treatment of dyes is very difficult to process as it remains a stable pollutant for a longer duration of time (Gupta et al., 2011a). Moreover, dyes are heat and light resistant and are also a nonbiodegradable pollutant. It hinders the growth of aquatic species and, hence, affects the photosynthesis process. For such emerging contaminants several techniques are available for removal: chemical oxidation process, the polyaniline (PANI)/ZnO nanocomposite system precipitation (Saravanan et al., 2016), carbon nanotube (Khani et al., 2010), photocatalytic degradation and electrochemical activity (Rajendran et al., 2016), photocatalytic treatment (Saravanan et al., 2013a, 2013b), facile thermal decomposition method (Saravanan et al., 2013f, 2015a), membrane filtration, photocatalytic degradation (Gupta et al., 2012), Fenton oxidation (Karthikeyan et al., 2012), ZnO/CdO nanocomposites (Saravanan et al., 2015b), ZnO/Ag/Mn₂O₃ nanocomposite (Saravanan et al., 2015c), ZnO/Ag/CdO nanocomposite (Saravanan et al., 2015d), catalyst ZnO/γ-Mn₂O₃ (Saravanan et al., 2014b), V₂O₅/ZnO nanocomposite system (Saravanan et al., 2014a), ZnO/Ag nanocomposite (Saravanan et al., 2013c), CeO₂/V₂O₅ and CeO₂/CuO catalysts (Saravanan et al., 2013d), ZnO/CuO nanocomposite (Saravanan et al., 2013e), and ion-exchange and adsorption process (Gupta et al., 1998; Mittal et al., 2009a, 2009b, 2010b; Pereira and Alves, 2012). Among all these technique adsorption processes is a very efficient technique for the removal of hazardous effluents like dyes from water bodies or water environment as it is low in operation cost, has minimum maintenance, simple design, resistance, and insensitivity toward harmful dyes (Mondal, 2008). An adsorbent with high porosity such as activated carbon (AC) compound is used for the removal of various industrial effluents (Gupta and Suhas, 2009). Many adsorbents such as activated blast furnace dust, sludge and slag from steel plants (Jain et al., 2007), mesoporous carbon material (Mohammadi et al., 2011), and AC (Asfaram et al., 2015) are used for the removal of dyes or textile effluents from water bodies.

One of the most commonly used cationic dyes in textile, paper, and pharmaceutical industries is safranin. Safranin dye without proper treatment is discharged into water bodies, and it poses threat to human health such as respiratory problem, skin irritation, and infections in the digestive system (Fayazi et al., 2015). Hence, it is very necessary to control such pollutant from mixing with water bodies to its minimum level or at lowest concentration level to reduce the harmful effects (Bhatnagar and Sillanpaa, 2010; Chowdhury et al., 2012; Saha and Mishra, 2012; Soni et al., 2012; Shakoor and Nasar, 2016).

In this research article, locally available B. tulda (a species of bamboo) was used as an adsorbent for the removal of safranin dye from water bodies.

B. tulda is available in abundance in nature, especially in South East Asian countries; it is low-cost bioadsorbent biodegradable in nature and readily not soluble in water. Cellulose and hemicelluloses are the main constituents of B. tulda. Some of the other elements are lignin, wax, etc. (Potsangbam et al., 2010; Singh et al., 2010; Wang and Yan, 2011). In comparison with other bioadsorbents like wood, tea leaves, and sawdust, the bamboo contains the higher presence of alkaline and silica which is suitable for adsorption techniques (Tomalang et al., 1980; Chen and Qin, 1985).

Raw lingo cellulose material like B. tulda gets biodegradable after a certain period of time, has low resistance, and less surface area and porosity (Wang et al., 2011). Therefore, to enhance the efficiency of B. tulda for effective dye removal, modification of B. tulda was done with sodium carbonate and HCI (Hydrochloric acid) for higher surface area and porosity.

The objective of the experiment is to find the effective and feasibly efficient low-cost bioadsorbent. From the initial experiment, it was found that sodium carbonate-modified B. tulda showed up as a better adsorbent compared with HCI-modified B. tulda and distilled washed B. tulda (DBT). Adsorption studies were done under effective operation parameters which include pH, dye concentration, adsorbent dosage, time of contact, temperature, and rotation per minute (RPM). Further studies were done with different isotherm models, kinetics studies, thermodynamics analysis with artificial neural network (ANN) model, regression analysis, and statistical error analysis. The characterization study of the adsorbent was done by SEM (scanning electron microscope), EDAX (energy diffractions and X-ray), and FTIR (Fourier transform infrared spectroscopy) analysis.

From the literature survey, removal of safranin dye using sodium carbonate modified B. tulda, HCI modified B. tulda, and distilled B. tulda has not been reviewed by any researcher. The physicochemical properties of the adsorbent have been analyzed. Moreover, there is no report of detailed characteristic study and modeling analysis by ANN and Regression Analysis (RA) Model Study along with Error analysis on modified B. tulda as an adsorbent in adsorption technique in the literature survey which established novelty of the research work.

Materials and Methods

Modification of bioadsorbent

B. tulda was collected from local market and made into small pieces of particle size 425–600 μm. It was washed with deionized water to remove any kind of foreign substances and then at 60°C it was oven dried. Some amount of DBT was then modified with 0.1 M Na₂CO₃ (sodium carbonate) and HCI (hydrochloric acid) separately for 6 h at rpm of 200. The excess of Na₂CO₃ and HCI onto B. tulda was later on washed properly to keep the pH at 7–7.5 and dried again at 60°C temperature and store it in a seal pack glass bottle for future experiment.

Adsorbate preparation

The chemical formula of safranin (SAF) is C₂₀H₁₉ N₄Cl, λ_max = 516 nm (Saha and Mishra, 2012; Fayazi et al., 2015), which was used as an adsorbate in the experiment. A stock solution of 1,000 mg/L was prepared with the required amount of safranin dye in distilled water. All experiments were done by diluting the stock solution with distilled water to attain the required concentration requirement.

Characterization analysis of B. tulda

SEM analysis was done to get high-resolution image of the adsorbent; it has a magnification power from 8× to 300,000× .

The FTIR analysis was done to identify the functional group active in the adsorbent for dye removal, ranges from 450 to 7,500 cm⁻¹. Infrared rays identify the functional groups present in the adsorbent.

Energy dispersive X-ray spectroscopy was done to identify the elements present in the adsorbent.

Brunauer–Emmett–Teller (BET) analysis was done to get the actual measurement of the surface enhancement of the modified bioadsorbent, before and after chemical modification.

Batch adsorption studies

Preliminary screening experiment

The preliminary screening experiment was studied with sodium carbonate treated B. tulda (NBT); hydrochloric acid treated B. tulda (HBT); and DBT. The optimum parameters were selected based on adsorption kinetic efficiency and maximum adsorption capacity of the adsorbent. Initial experiments were done for dye removal, and it was found that NBT is effective in safranin dye removal.

Batch kinetics studies

Batch studies were done with NBT and HBT adsorbent in 250 mL flasks with 50 mL total volume of safranin dye at 50 mg/L initial dye concentration with 10 g/L of dose added in solution and rotated at 200 rpm for 90 min. The operational effect of pH, dose, contact time, rpm, initial dye concentration, and temperature was investigated in the present research. At a certain interval, samples were analyzed in UV-VIS Spectrophotometer. The safranin adsorbed per unit of modified B. tulda in mg/g was calculated by the following Equation (1). \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} q = { \frac { ( { C_o } - { C_t } ) V } { \rm { m } } } , \tag { 1 } \end{align*} \end{document}

where q = amount of the adsorbate adsorbed at any time (mg/g). C_o = initial concentration of safranin in mg/L. C_t = equilibrium concentration of the dye in mg/L. V = volume of the solution in L, m = mass adsorbent used g. For correctness and accuracy, an average of three reading has been taken for analysis.

Adsorption isotherm studies

Adsorption isotherm studies were done to get the adsorption behavior and understand its bonding with adsorbent and adsorbate. It was conducted after obtaining the optimum parameters of rpm, temperature, pH, contact time, etc. by varying the dose from 2 g to 10 g/L with safranin (SAF) dye of 100 mg/L concentration. In this experiment Langmuir, Freundlich, Temkin, and Dubinin–Radushkevich isotherms were studied for the safranin dye removal from water bodies. Langmuir and Freundlich's isotherm is among the two most used isotherms by the researchers (Kumar and Acharya, 2013; Kumar and Kumar, 2015; Mohsenipour et al., 2015; Xiaopeng et al., 2016).

Adsorption kinetic model study

Kinetic analysis was done to determine the effectiveness of adsorption technique with respect to its time. The pseudo first order and pseudo second order kinetics models were calculated to get the rate constants. \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} log \left( { { q_e } - { q_t } } \right) = log \left( { { q_e } } \right) { \frac { { k_1 } } { 2.303 } } t , \end{align*} \end{document}

where q_e and q_t adsorbed dye mass (mg/g), t time, and K₁ is first order rate constant (/min)

Pseudo second order equation is: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} t / {q_t} \; = \;1 / {k_2}{q_2} \; + \left( {1 / {q_e}} \right) t , \end{align*} \end{document}

where k₂ is second order rate constant (g/[mg·min]) (Kumar and Kumar, 2015).

Activation energy

The minimum energy required by the bioadsorbate to get adsorbed is known as activation energy. The rate constant K at different temperatures was used to understand the activation energy using the Arrhenius equation: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \ln K = \ln A - { \frac { { E_a } } { RT } } , \end{align*} \end{document}

where E_a and A are Arrhenius Activation Energy and Arrhenius factor, respectively.

Thermodynamic analysis

The thermodynamic study was done to understand the temperature effect of the adsorbent. Gibbs free energy change (ΔG°), enthalpy change (ΔH°), and entropy change (ΔS°) were identified by van't Hoff equation (Kumar and Acharya, 2013; Kumar and Kumar, 2015). \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \Delta G^ \circ \; = \; - RT \ { \rm{ln}} \ K \end{align*} \end{document} \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \Delta G^ \circ \; = \; \Delta H^ \circ - T \Delta S^ \circ \end{align*} \end{document}

Study of ANN and RA

In this study MATLAB 8.1/R2013a, Neural Network Toolbox software was used to predict the adsorption effectiveness of safranin dye removal by NBT. Input variables for feed forward neural network were as follows: contact time (5–120 min), temperature (298–318 K), RPM (125–200), pH (3–10), initial concentration (50–200 mg/L), and dose (0.05–0.5 g). The percentage removal was taken as the output variable. All total 83 experimental data were taken, and among them, 65 were taken as training data and 18 as testing data; number of hidden nodes was 3, 4, and 5. The Levenberg–Marquardt (LM) back propagation algorithm and Bayesian Regularization (BR) algorithms were utilized. A statistical error such as coefficient of determination R², mean square error, mean absolute error, median absolute error, and root mean square error was used to evaluate the performance of the model. The regression analysis was done to predict the % removal of safranin dye by NBT; here % Removal (% R) has been taken as a dependent variable, whereas other parameters like pH, temperature, rpm, dose, initial concentration, and contact time are independent variables taken into consideration (Daneshvar et al., 2006; Behzat et al., 2011; Seniha et al., 2012; Ghaedi et al., 2014; Maghsoudi et al., 2015; Debnath and Dey, 2016).

Results and Discussion

SEM, EDAX, and FTIR analysis of B. tulda

Three SEM images were analyzed and shown in Fig. 1a–c for Distilled B. tulda (DBT), sodium bicarbonate treated B. tulda (NBT), and safranin dye adsorbed sodium bicarbonate B. tulda. After chemical modification of B. tulda with sodium bicarbonate, the pores get widen up compared to distilled washed B. tulda. Such modification helps in increase of pore cavity in adsorbent and results in the larger surface area for higher adsorption of safranin dye into it.

FIG. 1.

(a) SEM of DBT, (b) SEM of NBT, (c) SEM of safranin adsorbed NBT. DBT, distilled washed Bambusa tulda; NBT, sodium carbonate treated B. tulda; SEM, scanning electron microscope.

In Table 1 BET surface test was shown and that after treatment the surface area of NBT was found to be 0.6002 m²/g, which is double that of DBT. EDAX of safranin dye adsorbed NBT is shown in Fig. 2. From the EDAX analysis image, it is clear that after the safranin dye has been adsorbed by NBT, it attacks some elements like Na, Cl, etc. Similar results were found by Patil and Shrivastava (2015).

FIG. 2.

EDAX images of safranin adsorbed NBT. EDAX, energy diffractions and X-ray.

Table 1.

Number of Brunauer–Emmett–Teller Analysis

BET analysis	DBT	NBT
	0.3029 m²/g	0.6002 m²/g

BET, Brunauer–Emmett–Teller; DBT, distilled washed Bambusa tulda; NBT, sodium carbonate treated B. tulda.

FTIR analyses were done to identify the functional group and the molecular structure present in the B. tulda. The spectrum range of FTIR was from 4,000 to 400 cm⁻¹. FTIR image of safranin (SAF) dye adsorbed NBT is shown in Fig. 3. Peak shows were at 3476.26 cm⁻¹ (OH stretch, H-bonded), 2894.17 cm⁻¹ (CH stretch, strong), 1,639 cm⁻¹ (C = O stretch (strong) amide), 1,706 cm⁻¹ (C = O stretch and stretch (strong) aldehyde), 1,103 cm⁻¹ (C–O stretch, strong), and 519 cm⁻¹ (C–H bond). Adsorption of safranin dye from water environment occurs due to the presence of Hydroxyl and Carboxyl group in NBT; similar results were found in Zhang and Bing (2014); Reddy et al. (2017); and Kulkarni et al. (2017).

FIG. 3.

FTIR of sodium carbonate treated and safranin (SAF) dye adsorbed NBT. FTIR, Fourier transform infrared spectroscopy.

Screening experiments

Effect of contact time and initial concentration was analyzed and shown in Fig. 4. The figure represents the removal percentage of safranin dye and decrease in concentration level (Co mg/L) of safranin dye with respect to time for different adsorbents. The removal of safranin dye with respect to contact time from 0 to 90 min was examined. Three adsorbents were taken for analysis DBT, NBT, and HBT with an initial concentration of 50 mg/L, dose of 10 g/L, rpm 200, and pH 7. It was found that NBT adsorbent shows the maximum removal of safranin dye compared with HBT and DBT adsorbent. The removal rate was very high in the initial stage, and later on, after attaining the equilibrium time, there was no such significant removal of safranin dye. After 60 min there was removal of 98.75% safranin dye with initial concentration reduced to 2.32 from 50 mg/L when treated with NBT. The same observation was done with HBT showing 95.70% removal with initial concentration reduced to 8.00 mg/L, whereas DBT showed better results than HBT with 96.91% removal rate; the initial concentration reduced to 5.75 from 50 mg/L after 60 min.

FIG. 4.

Screening experiment by DBT, NBT, and HBT. HBT, hydrochloric acid treated B. tulda.

Operational parameter analysis

Contact time and initial concentration

The effect of contact time and initial concentration is analyzed in Fig. 5. The figure represents the removal percentage of safranin dye and decrease in concentration level (Co mg/L) of safranin dye with respect to time. Three concentrations of safranin dye were taken at 50, 100, and 200 mg/L. It was found that the 50 mg/L concentration has higher removal rate as the dye concentration increases in the active pores of the NBT adsorbent get saturated, and hence, no further adsorption takes place. Thus NBT provided the best removal rate because of more active sites on absorbent after chemical treatment. A similar observation was noticed by Kumar and Kumar (2015).

FIG. 5.

Effect of initial concentration of safranin by NBT.

Effect of pH

The pH test is one of the most important test analyses in adsorption technology. In this study the test analysis was done from pH 3 to 10. The Fig. 6 shows the percentage removal of safranin and concentration (Co mg/L) of safranin dye with respect to pH. In acidic zone (pH 3) there was less removal of safranin dye by NBT but the removal rate increased as pH increased and after pH 7 there was again slight decrease in removal rate. Two possible conclusion theories can be drawn from the observed pH test: (a) there is an electrostatic force of attraction between adsorbent and adsorbate and (b) the chemical interaction taking place between adsorbent and adsorbate. Moreover, the pH_ZPC value or zero point charge was found to be 7.6, which signifies that below 7.6 NBT charge is positive in nature and above is negative in nature. The negative charge of NBT gets attracted with the positive charge of the safranin dye and hence more adsorption takes place and results in higher dye removal rate. A similar observation was noticed by Hossein and Mohammadi (2015).

FIG. 6.

Effect of pH.

Effect of NBT dose

Effect of adsorbent dose on safranin dye removal was studied by varying the NBT dose from 1 to 10 g/L with an initial concentration of 50 mg/L, rpm 200, pH 7, and temperature 298 K, as shown in Fig. 7.

FIG. 7.

Effect of NBT dose.

The removal of safranin dye increased with increase in dosage, due to availability of more active pore cavity on the surface of NBT adsorbent. In the experiment, it was found that 10 g/L gives best removal rate of 98.22%. So, the optimum dose was taken as 10 g/L for the removal of safranin dye by NBT from water bodies. A similar observation was noticed by Etim et al. (2012); Rajurkar and Asha (2015); and Kumar and Kumar (2015).

Effect of temperature

Effect of temperature study was done to check the temperature variation between NBT adsorbent and safranin dye. In Fig. 8, the temperature varies from 298 to 318 K; the initial dye concentration was 50 mg/L, rpm 200, pH 7, and a dosage of 10 g/L. From the experiment, it was observed that there is an increase of removal rate at a higher temperature which indicates that the adsorption is endothermic in nature. The pores of NBT get increased and enhanced the adsorption of safranin dye. A similar observation was noticed by Maurya and Mittal (2011) and Khattri and Singh (2000).

FIG. 8.

Effect of temperature.

Effect of RPM

Another parameter for adsorption technique is RPM. In Fig. 9, it is seen that RPM experiments were done by varying the rotation speed from 125 to 250 rpm and initial concentration of 50 mg/L, pH 7, temperature 298 K, and dosage 10 g/L. As the rpm increases there is a higher removal rate but after 200 rpm there is a slight decrease in removal rate. It is due to an excess of development of centripetal forces which results in the creation of repulsive forces and as a result dye which was adsorbed in NBT gets repulsed a bit. Hence, 200 rpm was taken at an optimum level. A similar observation was noticed by Kumar and Kumar (2015) and Kumar and Bandyopadhyay (2006).

FIG. 9.

Effect of rpm.

Adsorption isotherm analysis

In Table 2, the experimental data were analyzed with four isotherm models Langmuir, Freundlich, Temkin, and Dubinin–Radushkevich isotherm. It was found that adsorption system best fits with Langmuir isotherm model compared to other models. The Q_max value was 32.26 mg/g with R² value 0.923. The K_F value was 7.834 mg/g with R² value 0.908. The K_T value was 1.319 L/mg with R² value 0.837, and q_e value was −2E–06 mg/g with R² value 0.661. The R_L value was 0.15 which symbolizes that the adsorption process is favorable in nature.

Table 2.

Adsorption Isotherm Study of Safranin onto Sodium Carbonate Treated B. tulda

Langmuir isotherms	Freundlich isotherms	Temkin isotherms	D-R isotherms
Q_max = 32.26 mg/g	K_F = 7.834 mg/g	K_T = 1.319 L/mg	q_s = −2E–06 mg/g
b = 0.113 L/mg R_L = 0.15	1/n = 0.284	B_T = 9.192	β = 19.10 m-mole²/J²
R² = 0.923	R² = 0.908	R² = 0.837	R² = 0.661

Adsorption kinetic model study

Kinetics analysis was done by putting the experiment values at different temperatures and initial dye concentration on pseudo first order and pseudo second order rate equation (Papita, 2012; Ertugay, 2014; Yang et al., 2014; Mehmood et al., 2015). All the results are shown in Tables 3 and 4. It was observed that the pseudo second order rate equation suits better than pseudo first order. At 50 mg/L of initial concentration, the R² value was 0.995 and K_p2 was 0.1253 with q_exp 5.0 mg/g and q_cal 5.025 mg/g, whereas at 308 K temperature the R² value was 0.999 and K_p2 was 0.1201 with q_exp 4.852 mg/g and q_cal 4.97 mg/g.

Table 3.

Adsorption Kinetic Parameters at Different Concentrations of Safranin

		Pseudo first order			Pseudo second order
C_o (mg/L)	q_e, Exp (mg/g)	q_e, Cal (mg/g)	kp₁	R²	q_e, Cal (mg/g)	kp₂	R²
50	5.0	1.0764	0.0460	0.924	5.025	0.1253	0.998
100	8.774	6.745	0.0437	0.943	9.7087	0.00464	0.966
200	13.054	10.789	0.0345	0.953	14.985	0.00429	0.991

Table 4.

Adsorption Kinetic Parameters at Different Temperatures of Safranin

		Pseudo first order			Pseudo second order
Temp (K)	q_e, Exp (mg/g)	q_e, Cal (mg/g)	kp₁	R²	q_e, Cal (mg/g)	kp₂	R²
298	4.86	2.013	0.898	0.928	4.985	0.1174	0.999
308	4.852	1.721	0.072	0.856	4.97	0.1201	0.999
318	4.759	1.162	0.0354	0.968	4.878	0.0751	0.998

Activation energy

Activation energy was 17.376 kJ/mol. From literature survey, it was found that if the value is in between 8.4 and 83.70 kJ/mol then it denotes that the adsorption system is chemical adsorption (Chan et al., 2006; Aljeboree et al., 2014; Kumar and Kumar, 2015; Ashfqaq et al., 2016).

Thermodynamic analysis

The thermodynamic analysis is explained in Table 5; it was found that at three different temperatures the value of ΔG° is negative in nature which denotes that the adsorption process is favorable and spontaneous in nature. ΔG° decreases with the increase of temperature, which shows the mobility of dye molecules at a higher temperature. Both ΔH° and ΔS° are positive in nature, so the process is endothermic in nature (Yakout and Elsherif, 2010; Chowdhury et al., 2011; Ramakant et al., 2017).

Table 5.

Thermodynamic Analysis at Different Temperatures of Safranin

Temp (K)	ΔG (kJ/mol)	ΔH (kJ/mol)	ΔS (J/mol)	R²
298	−7.99072
308	−8.29637	2.346	34	0.995
318	−8.68358

Comparative study of NBT versus AC

A comparative study was done for the removal of safranin dye from wastewater with NBT adsorbent and a commercially available AC compound. In Fig. 10, it was observed that after 1,440 min of treatment, NBT gave better results in comparison to AC for the removal of safranin dye. AC removed 91.40% of safranin dye, whereas NBT removed 98.28% of safranin dye. Thus it can be concluded that the NBT is a better adsorbent to be used for the removal of safranin dye. Performance level of various adsorbents including AC was discussed in Table 6.

FIG. 10.

Comparative study of AC and NBT. AC, activated carbon.

Table 6.

Performance Level of Various Adsorbents for the Removal of Safranin Dye

S. no	Adsorbent	Time (min)	pH	Initial concentration of dye (mg/L)	Rotation/ min	% Removal of safranin dye
1	NBT	60	7	50	200	98.15
2	HBT	60	7	50	200	95.70
3	DBT	60	7	50	200	96.91
4	Activated carbon	60	7	50	200	82.76

HBT, hydrochloric acid treated B. tulda.

ANN and RA Model Study

Modeling of adsorption studies was made by ANN and RA to understand the adsorption performance at different model analyses. Structure of ANN is shown in Fig. 11. All the experimental data were put up as per the condition used in ANN for both training and testing of the LM algorithm and BR algorithms. Fig. 12a–l shows the comparative study of experimental and predicted values of the LM and BR with hidden node 3, 4, and 5, and in Tables 7 and 8 it shows the training and testing data of ANN.

FIG. 11.

Structure of artificial neural network.

FIG. 12.

(a) LM algorithm with three hidden nodes for training data. (b) LM algorithm with three hidden nodes for testing data. (c) LM algorithm with four hidden nodes for training data. (d) LM algorithm with four hidden nodes for testing data. (e) LM algorithm with five hidden nodes for training data. (f) LM algorithm with five hidden nodes for testing data. (g) BR algorithm with three hidden nodes for training data. (h) BR algorithm with three hidden nodes for testing data. (i) BR algorithm with four hidden nodes for training data. (j) BR algorithm with four hidden nodes for testing data. (k) BR algorithm with five hidden nodes for training data. (l) BR algorithm with five hidden nodes for testing data. BR, Bayesian Regularization; LM, Levenberg–Marquardt.

Table 7.

Training Data of Artificial Neural Network Model

S. no.	Time (min)	Dose (g/L)	pH	Temp (K)	Inc Co (mg/L)	RPM	% Removal
1	5	0.5	7	298	100	200	45.75
2	5	0.5	7	298	200	200	24.93
3	5	0.5	7	298	50	200	77.14
4	5	0.5	7	318	50	200	74.32
5	10	0.5	7	298	50	200	81.84
6	10	0.5	7	298	200	200	51.27
7	10	0.5	7	298	50	200	81.84
8	10	0.5	7	308	50	200	82.54
9	10	0.5	7	318	50	200	78.48
10	15	0.5	7	298	50	200	85
11	15	0.5	7	298	200	200	59.92
12	15	0.5	7	298	50	200	85
13	15	0.5	7	318	50	200	83.68
14	20	0.5	7	298	50	200	93.1
15	20	0.5	7	298	100	200	86.54
16	20	0.5	7	298	200	200	64.77
17	20	0.5	7	298	50	200	93.1
18	20	0.5	7	318	50	200	84.88
19	30	0.5	7	298	50	200	95.68
20	30	0.5	7	298	100	200	90.22
21	30	0.5	7	298	200	200	73.27
22	30	0.5	7	298	50	200	95.68
23	30	0.5	7	308	50	200	95.74
24	30	0.5	7	318	50	200	85.74
25	45	0.5	7	298	50	200	96.36
26	45	0.5	7	298	100	200	92.05
27	45	0.5	7	298	200	200	79.94
28	45	0.5	7	298	50	200	96.36
29	45	0.5	7	318	50	200	88.72
30	60	0.1	7	298	100	200	64.32
31	60	0.15	7	298	100	200	76.65
32	60	0.2	7	298	100	200	78.85
33	60	0.3	7	298	100	200	91.96
34	60	0.35	7	298	100	200	92.91
35	60	0.4	7	298	100	200	94.48
36	60	0.45	7	298	100	200	96.48
37	60	0.05	7	298	100	200	53.63
38	60	0.1	7	298	100	200	63.28
39	60	0.15	7	298	100	200	94.14
40	60	0.2	7	298	100	200	96.69
41	60	0.3	7	298	100	200	98.58
42	60	0.35	7	298	100	200	98.77
43	60	0.4	7	298	100	200	98.83
44	60	0.45	7	298	100	200	98.93
45	60	0.5	7	298	100	200	99.23
46	60	0.5	3	298	100	200	88.2
47	60	0.5	6	298	100	200	95.6
48	60	0.5	8	298	100	200	98.4
49	60	0.5	10	298	100	200	98.4
50	60	0.5	7	298	50	125	86.49
51	60	0.5	7	298	50	175	89.29
52	60	0.5	7	298	50	225	89.26
53	60	0.5	7	298	50	200	96.7
54	60	0.5	7	298	100	200	93.58
55	60	0.5	7	298	200	200	83.62
56	60	0.5	7	298	50	200	96.7
57	60	0.5	7	308	50	200	96.54
58	60	0.5	7	318	50	200	93.86
59	90	0.5	7	298	50	200	97.22
60	90	0.5	7	298	100	200	95.88
61	90	0.5	7	298	50	200	97.22
62	90	0.5	7	308	50	200	97.24
63	90	0.5	7	318	50	200	95.18
64	120	0.5	7	298	50	200	98.48
65	120	0.5	7	298	100	200	96.79

RPM, rotation per minute.

Table 8.

Testing Data of Artificial Neural Network Model

S. no.	Time (min)	Dose (g/L)	pH	Temp (K)	Inc Co (mg/L)	RPM	% Removal
1	5	0.5	7	298	50	200	77.14
2	5	0.5	7	308	50	200	75.1
3	60	0.05	8	298	100	200	55.88
4	60	0.5	7	298	200	200	98.22
5	60	0.5	5	318	100	200	90.4
6	60	0.45	7	298	100	200	98.5
7	120	0.5	7	298	200	200	87.94
8	15	0.15	7	308	50	200	86.56
9	90	0.5	7	318	200	200	87.29
10	60	0.5	7	298	50	150	87.43
11	60	0.35	3	298	50	200	89.40
12	45	0.5	7	308	50	200	96.04
13	15	0.5	7	298	100	200	85.88
14	10	0.5	7	298	200	200	73.15
15	60	0.25	7	318	100	200	83.5
16	60	0.5	10	298	50	250	89.22
17	20	0.1	7	308	50	200	92.92
18	60	0.25	7	298	100	200	96.87

Statistical parameter analysis of ANN model is shown in Table 9. In Table 10, it was observed that two types of ANN algorithm were used. The LM with three hidden nodes has higher R² value with the lesser value of statistical error.

Table 9.

Statistical Error Analysis of Artificial Neural Network Model

			Statistical errors
Algorithms	No. of hidden nodes	R²	MAE	MSE	MEDAE	RMSE
LM	3	0.911	1.158	30.375	4.148	2.756
	4	0.880	1.479	69.133	4.517	4.157
	5	0.907	1.273	34.939	4.741	2.955
BR	3	0.660	1.851	172.674	4.292	6.570
	4	0.908	1.035	27.524	3.726	2.623
	5	0.891	1.129	33.542	4.169	2.896

BR, Bayesian Regularization; LM, Levenberg–Marquardt; MAE, mean absolute error; MEDAE, median absolute error; MSE, mean square error; RMSE, root mean square error.

Table 10.

Statistical Error Analysis of Artificial Neural Network Model

			Statistical errors
Algorithms	No. of hidden nodes	R²	MAE	MSE	MEDAE	RMSE
LM	3	0.911	1.158	30.375	4.148	2.756
	4	0.880	1.479	69.133	4.517	4.157
	5	0.907	1.273	34.939	4.741	2.955
BR	3	0.660	1.851	172.674	4.292	6.570
	4	0.908	1.035	27.524	3.726	2.623
	5	0.891	1.129	33.542	4.169	2.896

In Table 11, performance study of ANN versus RA model was done, and the ANN model was found to be better suitable compared to RA model. The analysis of variance (ANOVA) and the statistical analysis of variable in ANOVA are shown in Tables 12 and 13; RA model was analyzed with 95% confidence level. The p-value 4.27E–09 is very low, and the confidence level is almost 100%. It is found that the p-value of variable pH, temperature, and RPM is having a low confidence level, that is, (1 − p) < 0.95, and so the variables pH, temperature, and RPM are not signified for RA model analysis. However, p-value of time, dose, and initial concentration has high confidence level (1 − p) > 0.95.

Table 11.

Comparative Performance Study of Artificial Neural Network and Regression Analysis Model

			Statistical parameters
Model	Data	R²	MAE	MSE	MEDAE	RMSE
ANN	Training	0.963	2.156	1.167	1.434	3.820
	Testing	0.911	1.158	3.375	4.148	2.756
RA	Training	0.666	7.411	13.210	5.085	11.083
	Testing	0.425	2.001	12.181	5.972	5.504

Table 12.

Analysis of Variance

Source	df	SS	MS	F	Significance, p
Regression	6	7971.794	1328.632	12.70389	4.27E–09
Residual	58	6065.913	104.5847
Total	64	14037.71

Table 13.

Statistical Analysis of Variables in Analysis of Variance

Variable	Coefficients	Standard error	t Stat	p-value	Lower 95%	Upper 95%
Intercept	119.9751	67.90183	1.766891	0.082508	−15.9452	255.8954
Time	0.278024	0.048943	5.680607	4.55E–07	0.180055	0.375993
Dose	35.11522	11.28346	3.112098	0.002882	12.52893	57.70151
pH	1.330959	1.969189	0.675892	0.501795	−2.6108	5.27272
Temp (K)	−0.28006	0.208724	−1.34176	0.184902	−0.69787	0.137748
Inc Co	−0.15577	0.030455	−5.11487	3.7E–06	−0.21674	−0.09481
RPM	0.131199	0.125471	1.045653	0.30006	−0.11996	0.382357

In Table 13, 95% confidence level of lower and upper limits is shown, where pH, temperature, and RPM have zero value and cannot be considered for the t-test, whereas time, dose, and initial concentration can be considered for the t-test. Thus ANN model is better than RA model in the prediction of adsorption parameters within the ranges of the experimental data.

Chemical oxygen demand study and cost analysis

The chemical oxygen demand (COD) analysis was done to identify any difference in wastewater before and after treatment. It was found that COD level was 144 mg/L before treatment, and after treatment with NBT, it was 94 mg/L which is below the permissible limit of COD in wastewater. A similar observation was noticed by Kumar and Kumar (2015).

Currently, the cost of commercial AC is costly ($172.96/kg, Singh et al., 2016). The total cost of NBT, including drying, raw bamboo, and sodium carbonate, is Rs. 6.32 (approx.) per kg of NBT.

Conclusion

A detailed experimental study was done to remove the safranin dye from water bodies by NBT. The batch study experiment gave optimum results for safranin dye at pH 7, rpm 200, temperature 298 K, the initial dye concentration at 50 mg/L, and dose of 10 g/L with maximum adsorption capacity of 32.26 mg/g. The pseudo second order rate equation suits better than pseudo first order. At 50 mg/L of initial dye concentration, the R² value was 0.995 and K_p2 was 0.1253 with q_exp 5.0 mg/g and q_cal 5.025 mg/g. BET surface test was done and found that surface area of NBT after treatment is twice that of DBT. COD test was done, and it was found that the level of COD got reduced after treatment from 144 to 94 mg/L which is below the acceptable range of COD in wastewater. Comparing two adsorbent NBT and AC it was found that NBT (98.28% removal) shows better performance in safranin dye removal in comparison to AC (91.4% removal) at 1,440 min. In ANN the LM algorithm with three hidden nodes has higher R² value of 0.911 with lesser value of statistical error. Regression analysis model was analyzed with 95% confidence level. The p-value 4.27E–09 is very low, and the confidence level is almost 100%. It was found that the p-value of variable pH, temperature, and RPM is having a low confidence level, that is, (1 − p) < 0.95, and so the variables pH, temperature, and RPM are not signified for RA model analysis. However, p-value of time, dose, and initial concentration has high confidence level (1 − p) > 0.95. In the thermodynamic analysis, both ΔH° and ΔS° are positive in nature, so the process is endothermic in nature. The activation energy was found to be 17.376 kJ/mol. The total cost of NBT, including drying, raw bamboo, and sodium carbonate, is Rs. 6.32/kg of NBT. Taking consideration of all experiment results and model analysis, it is concluded that NBT is a good, low cost, and effective adsorbent for the removal of safranin dye from wastewater.

Footnotes

Acknowledgments

The authors thank Ministry of Human Resource, Government of India and TEQIP II NIT Silchar for providing financial and other support throughout the research and SAIF NEHU Shillong, SAIF IIT Bombay, CSIR Bhavnagar, and Pollution Control Board Silchar, Assam for technical help in SEM, EDAX, FTIR, BET, and COD analysis of work.

Author Disclosure Statement

No competing financial interests exist.

References

Aljeboree

, Abbas

, Ayad

, and Alkaim

(2014). Kinetics and equilibrium study for the adsorption of textile dyes on coconut shell activated carbon. Arab. J. Chem., 10, 2, 3381.

Asfaram

, Ghaedi

, Agarwal

, Tyagi

, and Gupta

V.K.

(2015). Removal of basic dye Auramine - O by ZnS: Cu nanoparticles loaded on activated carbon: Optimization of parameters using response surface methodology with central composite design. RSC Adv. 5, 18438.

Ashfqaq

, Gardazi

, Sherazi

, Ali

, Rashid

, and Bilal

(2016). Equilibrium kinetic and thermodynamic studies of Cr (VI) adsorption onto a novel adsorbent of Eucalyptus camaldulensis waste: Batch and column reactors. Korean J. Chem. Eng., 33, 10, 2898.

Behzat

, Olcayto

, and Mutlu

(2011). Use of BDST and an ANN model for prediction of dye adsorption efficiency of Eucalyptus camaldulensis barks in fixed-bed system. Expert Syst. Appl., 38, 949.

Bhatnagar

, and Sillanpaa

(2010). Utilization of agro-industrial and municipal waste materials as potential adsorbents for water treatment—A review. Chem. Eng. J., 157, 277.

Chan

W.H.

, Cheung

S.J.

, and McKay

(2006). Error analysis of adsorption isotherm models for acid dyes onto bamboo-derived activated carbon. Chinese J. Chem. Eng., 20, 535.

Chen

Y.D.

, Qin

W.L.

, Li

G.X.

, and Nimanna

(1985). The chemical composition of ten bamboo species. In Rao

A.N.

, et al., Eds. Recent Research on Bamboo. Proceedings of the International Bamboo Workshop, Hangzhou, China, October 6–14. Chinese Academy of Forestry, Beijing China. International Development Research Center, Ottawa, Canada, p. 110.

Chowdhury

, Mishra

, Kumar

, and Das

(2011). Adsorption thermodynamics, kinetics and isosteric heat of adsorption of malachite green onto chemically modified rice husk. Bioremed. J., 15, 2, 77.

Chowdhury

, Mishra

, Kushwah

, and Saha

(2012). Removal of safranin from aqueous solution by NaOH treated rice husk: Thermodynamic, kinetic and isosteric heat of adsorption. Asia-Pacific J. Chem. Eng., 7, 236.

10.

Daneshvar

, Khataee

A.R.

, and Djafarzadeh

(2006). The use of artificial neural networks (ANN) for modeling of decolorization of textile dye solution containing C. I. Basic Yellow 28 by electro coagulation process. J. Hazard. Mater. B, 137, 1788.

11.

Debnath

, and Dey

A.K.

(2016). Prediction of laboratory peak shear stress along the cohesive Soil-Geosynthetic interface using artificial neural network. Geotech. Geol. Eng., 35, 445.

12.

Ertugay

(2014). Adsorption, isotherm, kinetic & thermodynamic studies for MB by needles of Pinus sylvestris L. polish. J. Env. Stud., 23, 95.

13.

Etim

S.A.

, Umoren

, and Eduok

U.M.

(2012). Coconut coir dust as a low-cost adsorbent for the removal of cationic dye from aqueous solution. J. Saudi Chem. Soc., 20, 67.

14.

Fayazi

, Afzali

, Taher

M.A.

, Mostafavi

, and Gupta

V.K.

(2015). Removal of Safranin dye from aqueous solution using magnetic mesoporous clay: Optimization study. J. Mol. Liq., 212, 675.

15.

Ghaedi

, Zeinali

, Maghsoudi

, and Purkait

M.K.

(2014). Artificial Neural Network (ANN) Method for modeling of sunset yellow dye adsorption using nickel sulfide nanoparticle loaded on activated carbon: Kinetic and Isotherm Study. J. Dispers. Sci. Technol., 36, 1339.

16.

Gupta

V.K.

, Srivastava

S.K.

, Mohan

, and Sharma

(1998). Design parameters for fixed bed reactors of activated carbon developed from fertilizer waste for the removal of some heavy metal ions. Waste Manag. 17, 517.

17.

Gupta

V.K.

, and Suhas

(2009). Application of low-cost adsorbents for dye removal—A review. J. Environ. Manage., 90, 2313.

18.

Gupta

V.K.

, Jain

, Nayak

, Agarwal

, and Shrivastava

(2011a). Removal of the hazardous dye—Tartrazine by photodegradation on titanium dioxide surface. Mat. Sci. Eng. C., 31, 1062.

19.

Gupta

V.K.

, Jain

, Mittal

, Saleh

T.A.

, Nayak

, Agarwal

, and Sikarwar

(2012). Photo-catalytic degradation of toxic dye amaranth on TiO2/UV in aqueous suspensions. Mater. Sci. Eng. C., 32, 12.

20.

Hossein

, and Mohammadi

(2015). Quince seed mucilage magnetic nanocomposites as novel bioadsorbents for efficient removal of cationic dyes from aqueous solutions, Carbohydr. Polym. 134, 213.

21.

Jain

A.K.

, Gupta

V.K.

, Bhatnagar

, and Suhas . (2007). A comparative study of adsorbents prepared from industrial wastes for removal of dyes. J. Sep. Sci. Technol., 38, 2.

22.

Karthikeyan

, Gupta

V.K.

, Boopathy

, Titus

, and Sekaran

(2012). A new approach for the degradation of high concentration of aromatic amine by heterocatalytic Fenton oxidation: Kinetic and spectroscopic studies. J. Mol. Liq., 173, 153.

23.

Khani

, Rofouei

M.K.

, Arab

, Gupta

V.K.

, and Vafaei

(2010). Multi-walled carbon nanotubes-ionic liquid-carbon paste electrode as a super selectivity sensor: Application to potentiometric monitoring of mercury ion (II). J. Hazard. Mater., 183, 402.

24.

Khattri

, and Singh

(2000). Colour removal of synthetic dye wastewater using bioadsorbent. Water Air Pollut. 120, 283.

25.

Kulkarni

M.R.

, Revanth

, Acharya

, and Bhat

(2017). Removal of crystal violet dye from aqueous solution using water hyacinth: Equilibrium, kinetics and thermodynamics study. Resour.-Eff. Technol., 3, 71.

26.

Kumar

, and Kumar

(2015). Adsorption of malachite green in aqueous solution onto sodium carbonate treated rice hush. Korean J. Chem. Eng., 32, 1655.

27.

Kumar

, and Acharya

(2013). Fixed Bed Column Study for the removal of lead from aquatic environment by NCRH. Res. J. Recent Sci., 2, 9.

28.

Kumar

, and Bandyapadhya

(2006). Sorption of cadmium from aqueous solution using pretreated rice husk. Bioresour. Technol., 97, 104.

29.

Maghsoudi

, Ghaedi

, Zinali

, Ghaedi

A.M.

, and Habibi

M.H.

(2015). Artificial neural network (ANN) method for modeling of sunset yellow dye adsorption using zinc oxide nanorods loaded on activated carbon: Kinetic and isotherm study. Spectrochim. Acta A, 134, 1.

30.

Maurya

N.S.

, and Mittal

A.K.

(2011). Biosorptive uptake of cationic dyes from aqueous phase using immobilised dead macro fungal biomass. Int. J. Environ. Technol. Manag., 14, 282.

31.

Mehmood

, Sheher

, and Aisha

(2015). Effective removal of crystal violet and eosin B from aqueous solution using syzygium cumini leaves: A comparative study of acidic and basic dye on single adsorbent. Korean J. Chem. Eng., 32, 882.

32.

Mittal

, Mittal

, Malviya

, and Gupta

V.K.

(2009a). Adsorptive removal of hazardous anionic dye Congo red from wastewater using waste materials and recovery by desorption. J. Colloid Interf. Sci., 340, 1.

33.

Mittal

, Mittal

, Malviya

, and Gupta

V.K.

(2009b). Adsorptive removal of hazardous anionic dye “Congo red” from wastewater using waste materials and recovery by desorption. J. Colloid Interf. Sci., 340, 16.

34.

Mittal

, Mittal

, Malviya

, and Gupta

V.K.

(2010a). Removal and recovery of Chrysoidine Y from aqueous solutions by waste materials. J. Colloid Interf. Sci., 344, 497.

35.

Mittal

, Mittal

, Malviya

, Kaura

, and Gupta

V.K.

(2010b). Decoloration treatment of a hazardous triaryl methane dye, Light Green SF (Yellowish) by waste material adsorbents. J. Colloid Interf. Sci., 342, 518.

36.

Mohammadi

, Khani

, Gupta

V.K.

, Amereh

, and Agarwal

(2011). Adsorption process of methyl orange dye onto mesoporous carbon material–kinetic and thermodynamic studies. J. Colloid Interf. Sci., 362, 457.

37.

Mohsenipour

, Shahid

, and Ebrahimi

(2015). Nitrate adsorption on clay kaolin: Batch tests. J. Chem., 397069, 7.

38.

Mondal

(2008). Methods of dye removal from dye house effluent—An overview. Environ. Eng. Sci., 25, 383.

39.

Papita

, Chakroborty

, and Chowdhury

(2012). Batch and continuous (fixed-bed column) biosorption of crystal violet by Artocarpus heterophyllus (jackfruit) leaf powder. Colloids Surf. B Biointerfaces, 92, 262.

40.

Patil

M.R.

, and Shrivastava

V. S.

(2015). Adsorption removal of carcinogenic acid violet 19 dye from aqueous solution by polyaniline-Fe₂O₃ magnetic nano-composite. J. Mater. Environ. Sci., 6, 11.

41.

Pereira

, and Alves

(2012). Dye-environmental impact and remediation in: Environmental Protection Strategies for Sustainable Development. Strategies for Sustainability, Springer Science + Business Media B.V., Dordrecht, 111. doi.org/10.1007/978-94-007-1591-2_4.

42.

Potsangbam

K.S.

, Sorokhaibam

P.D.

, Keisham

, Devi , Deshworjit

S.N.

, and Pinokiyo

(2010). Bambusa tulda Roxb. In Manipur State, India: Exploring the Local Values and Commercial Implications. Notulae Scientia Biologicae, 2, 35. ISSN 2067-3205.

43.

Rajendran

, Khan

M.M.

, Gracia

, Qin

, Gupta

V.K.

, and Arumainathan

(2016). Ce3+-ion-induced visible-light photocatalytic degradation and electrochemical activity of ZnO/CeO₂ nanocomposite. Sci. Rep., 6, 31641.

44.

Rajurkar

, and Asha

(2015). Removal of CV from water using Chitosan and Saraca indica leaves. J. Appl. Chem., 4, 1446.

45.

Ramakant

S.I.

, Dilip

H.L

and Prashant

T.D.

(2017). Adsorption of phenol onto banana peels activated carbon. KSCE J. Civil Eng., 21, 100.

46.

Reddy

M.C

, Nirmala

, and Ashwini

(2017). Bengal Gram Seed Husk as an adsorbent for the removal of dye from aqueous solutions—Batch studies. Arab. J. Chem., 10, 2554.

47.

Saha

, and Mishra

(2012). Adsorption of safranin onto chemically modified rice husk in a upward packed bed reactor: ANN modeling. Pollution, 44, 7579.

48.

Saravanan

, Karthikeyan

, Gupta

V.K.

, Thirumal

, Thangadurai

, Narayanan

, and Stephen

(2013a). ZnO/Ag nanocomposite: An efficient catalyst for degradation studies of textile effluents under visible light. Mater. Sci. Eng. C, 33, 2235.

49.

Saravanan

, Thirumal

, Gupta

V.K.

, Narayanan

, and Stephen

(2013b). The photocatalytic activity of ZnO prepared by simple thermal decomposition method at various temperatures. J. Mol. Liq., 177, 394.

50.

Saravanan

, Gupta

V.K.

, Narayanan

, and Stephen

(2013c). Comparative study on photocatalytic activity of ZnO prepared by different methods. J. Mol. Liq., 181, 133.

51.

Saravanan

, Joicy

, Gupta

V.K.

, Narayanan

, and Stephen

(2013d). Visible light induced degradation of methylene blue using CeO₂/V₂O₅ and CeO₂/CuO catalysts. Mater. Sci. Eng. C, 33, 4725.

52.

Saravanan

, Karthikeyan

, Gupta

V.K.

, Sekaran

, Narayanan

, and Stephen

(2013e). Enhanced photocatalytic activity of ZnO/CuO nanocomposite for the degradation of textile dye on visible light illumination. Mater. Sci. Eng. C, 33, 91.

53.

Saravanan

, Gupta

V.K.

, Prakash

, Narayanan

, and Stephen

(2013f). Synthesis, characterization and photocatalytic activity of novel Hg doped ZnO nanorods prepared by thermal decomposition method. J. Mol. Liq., 178, 88.

54.

Saravanan

, Gupta

V.K.

, Mosquerae

, and Graciaa

(2014a). Preparation and characterization of V₂O₅/ZnO nanocomposite system for photocatalytic application. J. Mol. Liq., 198, 409.

55.

Saravanan

, Gupta

V.K.

, Narayanan

, Stephen

(2014b). Visible light degradation of textile effluent using novel catalyst ZnO/γ-Mn₂O₃. J. Taiwan Inst. Chem. Eng., 45, 1910.

56.

Saravanan

, Gupta

V.K.

, Mosquerae

, Graciaa

, Narayanan

, and Stephen

(2015a). Visible light induced degradation of methyl orange using β-Ag0.333V2O5 nanorod catalysts by facile thermal decomposition method. J. Saudi Chem. Soc., 19, 521.

57.

Saravanan

, Gracia

, Khan

M.M.

, Poornima

, Gupta

V.K.

, Narayanan

, and Stephen

(2015b). ZnO/CdO nanocomposites for textile effluent degradation and electrochemical detection. J. Mol. Liq., 209, 374.

58.

Saravanan

, Khan

M.M.

, Gupta

V.K.

, Mosquerae

, Graciaa

, Narayanan

, and Stephen

(2015c). ZnO/Ag/Mn₂O₃ nanocomposite for visible light-induced industrial textile effluent degradation, uric acid and ascorbic acid sensing and antimicrobial activity. RSC Adv. 5, 34645.

59.

Saravanan

, Khan

M.M.

, Gupta

V.K.

, Mosquerae

, Graciaa

, Narayanan

, and Stephen

(2015d). ZnO/Ag/CdO nanocomposite for visible light-induced photocatalytic degradation of industrial textile effluents. J. Colloid Interf. Sci., 452, 126.

60.

Saravanan

, Sacari

, Gracia

, Khan

M. M.

, Mosquera

, and Gupta

V.K.

(2016). Conducting PANI stimulated ZnO system for visible light photocatalytic degradation of coloured dyes. J. Mol. Liq., 221, 1029.

61.

Seniha

, Emriye

P.A. K.

, and Saadet

(2012). Modeling the adsorption of textile dye on organo clay using an artificial neural network. Dyes Pigm. 95, 102.

62.

Shakoor

, and Nasar

(2016). Removal of methylene blue dye from artificially contaminated water using citrus limetta peel waste as a very low cost adsorbent. J Taiwan Inst. Chem. Eng., 66, 154.

63.

Singh

, Chandra

, and Gautam

(2016). Development of inexpensive adsorbent from agro-waste for phenol adsorption. J. Sci. Ind. Res., 75, 444.

64.

Singh

P.S.

, Devi

S.P.

, Devi

K.K.

, Ningombam

D.S.

, and Athokpam

(2010). Bambusa tulda Roxb. in Manipur State, India: Exploring the local values and commercial implications. Notulae Sci. Biol., 2, 35.

65.

Soni

, Sharma

A.K.

, Srivastava

J.K.

, and Yadav

J.S.

(2012). Adsorptive removal of Methylene blue dye from an aqueous solution using water hyacinth root powder as a low cost adsorbent. Int. J. Chem. Sci. Appl., 3, 338.

66.

Tomalang

F.N.

, Lopez

A.R.

, Semara

J.A.

, Casin

R.F.

, and Espiloy

Z.B.

(1980). Properties and utilization of Philippine erect bamboo. In Lessard

and Chouinard

, Eds. International Seminar on Bamboo Research in Asia. Singapore, May 28–30. International Development Research Center and the International Union of Forestry Research Organization, Singapore, p. 266.

67.

Wang

, and Yan

(2011). Adsorptive removal of direct yellow from the aqueous solution from bamboo charcoal activated with chemicals. Desalination, 271, 81.

68.

Xiaopeng

, Fei

, Long

, Hong

, Guangcai

, Fasheng

, and Liping

(2016). Adsorption of Levofloxacin to Goethite: Batch and column studies. Environ. Eng. Sci., 33. Doi:10.1089/ees.2015.0379

69.

Yakout

S.M.

, and Elsherif

(2010). Batch kinetics, isotherm and thermodynamic studies of adsorption of strontium from aqueous solutions onto low-cost rice-straw based carbons. Carbon—Sci. Tech., 1, 144.

70.

Yang

, Lin

, Wei

, and Wang

(2014). Evaluation of adsorption potential of bamboo bio char for metal-complex dye: Equilibrium, kinetics and artificial neural network modeling. Inter. J. Enviro. Sci. Tech., 11, 1093.

71.

Zhang

, and Bing

(2014). Removal of methylene blue from aqueous solutions by chemically modified bamboo. Chemosphere, 111, 225.